localdocs: implement .docx support

.gitmodules

@@ -11,3 +11,6 @@
 [submodule "gpt4all-chat/deps/fmt"]
 	path = gpt4all-chat/deps/fmt
 	url = https://github.com/fmtlib/fmt.git
+[submodule "gpt4all-chat/deps/DuckX"]
+	path = gpt4all-chat/deps/DuckX
+	url = https://github.com/nomic-ai/DuckX.git

gpt4all-chat/CHANGELOG.md

@@ -8,6 +8,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ### Added
 - Add bm25 hybrid search to localdocs ([#2969](https://github.com/nomic-ai/gpt4all/pull/2969))
+- LocalDocs support for .docx files ([#2986](https://github.com/nomic-ai/gpt4all/pull/2986))
 
 ### Changed
 - Rebase llama.cpp on latest upstream as of September 26th ([#2998](https://github.com/nomic-ai/gpt4all/pull/2998))

gpt4all-chat/CMakeLists.txt

@@ -86,14 +86,9 @@ get_filename_component(Qt6_ROOT_DIR "${Qt6_ROOT_DIR}/.." ABSOLUTE)
 message(STATUS "qmake binary: ${QMAKE_EXECUTABLE}")
 message(STATUS "Qt 6 root directory: ${Qt6_ROOT_DIR}")
 
-set (CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
-
-set(FMT_INSTALL OFF)
-set(BUILD_SHARED_LIBS_SAVED "${BUILD_SHARED_LIBS}")
-set(BUILD_SHARED_LIBS OFF)
-add_subdirectory(deps/fmt)
-set(BUILD_SHARED_LIBS "${BUILD_SHARED_LIBS_SAVED}")
+set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
 
+add_subdirectory(deps)
 add_subdirectory(../gpt4all-backend llmodel)
 
 set(CHAT_EXE_RESOURCES)
@@ -133,9 +128,6 @@ if (APPLE)
     list(APPEND CHAT_EXE_RESOURCES "${LOCAL_EMBEDDING_MODEL_PATH}")
 endif()
 
-set(QAPPLICATION_CLASS QGuiApplication)
-add_subdirectory(deps/SingleApplication)
-
 if (DEFINED GGML_METALLIB)
     set_source_files_properties("${GGML_METALLIB}" PROPERTIES GENERATED ON)
 endif()
@@ -335,7 +327,7 @@ target_include_directories(chat PRIVATE deps/usearch/include
 target_link_libraries(chat
     PRIVATE Qt6::Core Qt6::HttpServer Qt6::Pdf Qt6::Quick Qt6::Sql Qt6::Svg)
 target_link_libraries(chat
-    PRIVATE llmodel SingleApplication fmt::fmt)
+    PRIVATE llmodel SingleApplication fmt::fmt duckx::duckx)
 
 
 # -- install --

gpt4all-chat/deps/CMakeLists.txt

@@ -0,0 +1,10 @@
+set(BUILD_SHARED_LIBS OFF)
+
+set(FMT_INSTALL OFF)
+add_subdirectory(fmt)
+
+set(QAPPLICATION_CLASS QGuiApplication)
+add_subdirectory(SingleApplication)
+
+set(DUCKX_INSTALL OFF)
+add_subdirectory(DuckX)

gpt4all-chat/qml/LocalDocsSettings.qml

@@ -70,7 +70,7 @@ MySettingsTab {
                     /* Blacklist common unsupported file extensions. We only support plain text and PDFs, and although we
                      * reject binary data, we don't want to waste time trying to index files that we don't support. */
                     exts = exts.filter(e => ![
-                        /* Microsoft documents  */ "rtf", "docx", "ppt", "pptx", "xls", "xlsx",
+                        /* Microsoft documents  */ "rtf", "ppt", "pptx", "xls", "xlsx",
                         /* OpenOffice           */ "odt", "ods", "odp", "odg",
                         /* photos               */ "jpg", "jpeg", "png", "gif", "bmp", "tif", "tiff", "webp",
                         /* audio                */ "mp3", "wma", "m4a", "wav", "flac",